/[Biblio-Isis]/trunk/lib/Biblio/Isis.pm
This is repository of my old source code which isn't updated any more. Go to git.rot13.org for current projects!
ViewVC logotype

Diff of /trunk/lib/Biblio/Isis.pm

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 3 by dpavlin, Tue Dec 28 01:48:44 2004 UTC revision 25 by dpavlin, Fri Dec 31 05:43:20 2004 UTC
# Line 2  package IsisDB; Line 2  package IsisDB;
2  use strict;  use strict;
3    
4  use Carp;  use Carp;
5    use File::Glob qw(:globally :nocase);
6    
7  use Data::Dumper;  use Data::Dumper;
8    
9  BEGIN {  BEGIN {
10          use Exporter ();          use Exporter ();
11          use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);          use vars qw ($VERSION @ISA @EXPORT @EXPORT_OK %EXPORT_TAGS);
12          $VERSION     = 0.01;          $VERSION     = 0.08;
13          @ISA         = qw (Exporter);          @ISA         = qw (Exporter);
14          #Give a hoot don't pollute, do not export more than needed by default          #Give a hoot don't pollute, do not export more than needed by default
15          @EXPORT      = qw ();          @EXPORT      = qw ();
# Line 18  BEGIN { Line 20  BEGIN {
20    
21  =head1 NAME  =head1 NAME
22    
23  IsisDB - Read CDS/ISIS database  IsisDB - Read CDS/ISIS, WinISIS and IsisMarc database
24    
25  =head1 SYNOPSIS  =head1 SYNOPSIS
26    
27    use IsisDB    use IsisDB;
28    
29    my $isis = new IsisDB(    my $isis = new IsisDB(
30          isisdb => './cds/cds',          isisdb => './cds/cds',
31    );    );
32    
33      for(my $mfn = 1; $mfn <= $isis->{'maxmfn'}; $mfn++) {
34            print $isis->to_ascii($mfn),"\n";
35      }
36    
37  =head1 DESCRIPTION  =head1 DESCRIPTION
38    
39  This module will read CDS/ISIS databases and create hash values out of it.  This module will read ISIS databases created by DOS CDS/ISIS, WinIsis or
40  It can be used as perl-only alternative to OpenIsis module.  IsisMarc. It can be used as perl-only alternative to OpenIsis module.
41    
42    It can create hash values from data in ISIS database (using C<to_hash>),
43    ASCII dump (using C<to_ascii>) or just hash with field names and packed
44    values (like C<^asomething^belse>).
45    
46    Unique feature of this module is ability to C<include_deleted> records.
47    It will also skip zero sized fields (OpenIsis has a bug in XS bindings, so
48    fields which are zero sized will be filled with random junk from memory).
49    
50    It also has support for identifiers (only if ISIS database is created by
51    IsisMarc), see C<to_hash>.
52    
53    This will module will always be slower than OpenIsis module which use C
54    library. However, since it's written in perl, it's platform independent (so
55    you don't need C compiler), and can be easily modified. I hope that it
56    creates data structures which are easier to use than ones created by
57    OpenIsis, so reduced time in other parts of the code should compensate for
58    slower performance of this module (speed of reading ISIS database is
59    rarely an issue).
60    
61  =head1 METHODS  =head1 METHODS
62    
# Line 50  It can be used as perl-only alternative Line 76  It can be used as perl-only alternative
76  # some binary reads  # some binary reads
77  #  #
78    
 sub Read32 {  
         my $self = shift;  
   
         my $f = shift || die "Read32 needs file handle";  
         read($$f,$b,4) || die "can't read 4 bytes from $$f from position ".tell($f);  
         return unpack("l",$b);  
 }  
   
79  =head2 new  =head2 new
80    
81  Open CDS/ISIS database  Open ISIS database
82    
83   my $isis = new IsisDB(   my $isis = new IsisDB(
84          isisdb => './cds/cds',          isisdb => './cds/cds',
85          read_fdt => 1,          read_fdt => 1,
86            include_deleted => 1,
87            hash_filter => sub {
88                    my $v = shift;
89                    $v =~ s#foo#bar#g;
90            },
91          debug => 1,          debug => 1,
92   );   );
93    
# Line 74  Options are described below: Line 97  Options are described below:
97    
98  =item isisdb  =item isisdb
99    
100  Prefix path to CDS/ISIS. It should contain full or relative path to database  This is full or relative path to ISIS database files which include
101  and common prefix of C<.FDT>, C<.MST>, C<.CNT>, C<.XRF> and C<.MST> files.  common prefix of C<.MST>, and C<.XRF> and optionally C<.FDT> (if using
102    C<read_fdt> option) files.
103    
104    In this example it uses C<./cds/cds.MST> and related files.
105    
106  =item read_fdt  =item read_fdt
107    
108  Boolean flag to specify if field definition table should be read. It's off  Boolean flag to specify if field definition table should be read. It's off
109  by default.  by default.
110    
111    =item include_deleted
112    
113    Don't skip logically deleted records in ISIS.
114    
115    =item hash_filter
116    
117    Filter code ref which will be used before data is converted to hash.
118    
119  =item debug  =item debug
120    
121  Dump a C<lot> of debugging output.  Dump a B<lot> of debugging output.
122    
123  =back  =back
124    
# Line 97  sub new { Line 131  sub new {
131          my $self = {};          my $self = {};
132          bless($self, $class);          bless($self, $class);
133    
134          $self->{isisdb} = {@_}->{isisdb} || croak "new needs database name as argument!";          croak "new needs database name (isisdb) as argument!" unless ({@_}->{isisdb});
135    
136          $self->{debug} = {@_}->{debug};          foreach my $v (qw{isisdb debug include_deleted hash_filter}) {
137                    $self->{$v} = {@_}->{$v};
138            }
139    
140            my @isis_files = grep(/\.(FDT|MST|XRF|CNT)$/i,glob($self->{isisdb}."*"));
141    
142            foreach my $f (@isis_files) {
143                    my $ext = $1 if ($f =~ m/\.(\w\w\w)$/);
144                    $self->{lc($ext)."_file"} = $f;
145            }
146    
147            my @must_exist = qw(mst xrf);
148            push @must_exist, "fdt" if ($self->{read_fdt});
149    
150            foreach my $ext (@must_exist) {
151                    croak "missing ",uc($ext)," file in ",$self->{isisdb} unless ($self->{$ext."_file"});
152            }
153    
154            print STDERR "## using files: ",join(" ",@isis_files),"\n" if ($self->{debug});
155    
156          # if you want to read .FDT file use read_fdt argument when creating class!          # if you want to read .FDT file use read_fdt argument when creating class!
157          if ({@_}->{read_fdt} && -e $self->{isisdb}.".FDT") {          if ($self->{read_fdt} && -e $self->{fdt_file}) {
158    
159                  # read the $db.FDT file for tags                  # read the $db.FDT file for tags
160                  my $fieldzone=0;                  my $fieldzone=0;
161    
162                  open(fileFDT, $self->{isisdb}.".FDT") || croak "can't read '$self->{isisdb}.FDT': $!";                  open(fileFDT, $self->{fdt_file}) || croak "can't read '$self->{fdt_file}': $!";
163    
164                  while (<fileFDT>) {                  while (<fileFDT>) {
165                          chomp;                          chomp;
# Line 131  sub new { Line 183  sub new {
183    
184          # Get the Maximum MFN from $db.MST          # Get the Maximum MFN from $db.MST
185    
186          open(fileMST,$self->{isisdb}.".MST") || croak "can't read '$self->{isisdb}.MST': $!";          open($self->{'fileMST'}, $self->{mst_file}) || croak "can't open '$self->{mst_file}': $!";
187    
188          # MST format:   (* = 32 bit signed)          # MST format:   (* = 32 bit signed)
189          # CTLMFN*       always 0          # CTLMFN*       always 0
# Line 139  sub new { Line 191  sub new {
191          # NXTMFB*       last block allocated to master file          # NXTMFB*       last block allocated to master file
192          # NXTMFP        offset to next available position in last block          # NXTMFP        offset to next available position in last block
193          # MFTYPE        always 0 for user db file (1 for system)          # MFTYPE        always 0 for user db file (1 for system)
194          seek(fileMST,4,0);          seek($self->{'fileMST'},4,0);
195          $self->{'NXTMFN'}=$self->Read32(\*fileMST) || carp "NXTNFN is zero";  
196            my $buff;
197    
198            read($self->{'fileMST'}, $buff, 4);
199            $self->{'NXTMFN'}=unpack("l",$buff) || carp "NXTNFN is zero";
200    
201          # save maximum MFN          # save maximum MFN
202          $self->{'maxmfn'} = $self->{'NXTMFN'} - 1;          $self->{'maxmfn'} = $self->{'NXTMFN'} - 1;
203    
204          close(fileMST);  
205    
206    
207            print STDERR Dumper($self),"\n" if ($self->{debug});
208    
209            # open files for later
210            open($self->{'fileXRF'}, $self->{xrf_file}) || croak "can't open '$self->{xrf_file}': $!";
211    
212            $self ? return $self : return undef;
213    }
214    
215    =head2 read_cnt
216    
217    This function is not really used by module, but can be useful to find info
218    about your index (if debugging it for example).
219    
220      print Dumper($isis->read_cnt);
221    
222    =cut
223    
224    sub read_cnt  {
225            my $self = shift;
226    
227            croak "missing CNT file in ",$self->{isisdb} unless ($self->{cnt_file});
228    
229          # Get the index information from $db.CNT          # Get the index information from $db.CNT
230        
231          open(fileCNT, $self->{isisdb}.".CNT") || croak "can't read '$self->{isisdb}.CNT': $!";          open(fileCNT, $self->{cnt_file}) || croak "can't read '$self->{cnt_file}': $!";
232    
233          # There is two 26 Bytes fixed lenght records          # There is two 26 Bytes fixed lenght records
234    
# Line 173  sub new { Line 252  sub new {
252                  my $buff = shift || return;                  my $buff = shift || return;
253                  my @arr = unpack("ssssssllls", $buff);                  my @arr = unpack("ssssssllls", $buff);
254    
255                    print STDERR "unpack_cnt: ",join(" ",@arr),"\n" if ($self->{'debug'});
256    
257                  my $IDTYPE = shift @arr;                  my $IDTYPE = shift @arr;
258                  foreach (@flds) {                  foreach (@flds) {
259                          $self->{$IDTYPE}->{$_} = abs(shift @arr);                          $self->{cnt}->{$IDTYPE}->{$_} = abs(shift @arr);
260                  }                  }
261          }          }
262    
263          my $buff;          my $buff;
264    
265          read(fileCNT, $buff, 26);          read(fileCNT, $buff, 26);
266          $self->unpack_cnt($buff);          $self->unpack_cnt($buff);
267    
268          read(fileCNT, $buff, 26);          read(fileCNT, $buff, 26);
269          $self->unpack_cnt($buff);          $self->unpack_cnt($buff);
270    
   
271          close(fileCNT);          close(fileCNT);
272    
273          print Dumper($self) if ($self->{debug});          return $self->{cnt};
   
         $self ? return $self : return undef;  
274  }  }
275    
276  =head2 GetMFN  =head2 fetch
277    
278  Read record with selected MFN  Read record with selected MFN
279    
280    my $rec = $isis->GetMFN(55);    my $rec = $isis->fetch(55);
281    
282  Returns hash with keys which are field names and values are unpacked values  Returns hash with keys which are field names and values are unpacked values
283  for that field.  for that field like this:
284    
285      $rec = {
286        '210' => [ '^aNew York^cNew York University press^dcop. 1988' ],
287        '990' => [ '2140', '88', 'HAY' ],
288      };
289    
290  =cut  =cut
291    
292  sub GetMFN {  sub fetch {
293          my $self = shift;          my $self = shift;
294    
295          my $mfn = shift || croak "GetMFN needs MFN as argument!";          my $mfn = shift || croak "fetch needs MFN as argument!";
296    
297          print "GetMFN: $mfn\n" if ($self->{debug});          # is mfn allready in memory?
298            my $old_mfn = $self->{'current_mfn'} || -1;
299            return $self->{record} if ($mfn == $old_mfn);
300    
301          open(fileXRF, $self->{isisdb}.".XRF") || croak "can't open '$self->{isisdb}.XRF': $!";          print STDERR "## fetch: $mfn\n" if ($self->{debug});
302    
303          # XXX check this?          # XXX check this?
304          my $mfnpos=($mfn+int(($mfn-1)/127))*4;          my $mfnpos=($mfn+int(($mfn-1)/127))*4;
305    
306          print "seeking to $mfnpos in file '$self->{isisdb}.XRF'\n" if ($self->{debug});          print STDERR "## seeking to $mfnpos in file '$self->{xrf_file}'\n" if ($self->{debug});
307          seek(fileXRF,$mfnpos,0);          seek($self->{'fileXRF'},$mfnpos,0);
308    
309          # read XRFMFB abd XRFMFP          my $buff;
         my $pointer=$self->Read32(\*fileXRF);  
310    
311          my $XRFMFB = int($pointer/2048);          # delete old record
312          my $XRFMFP = $pointer - ($XRFMFB*2048);          delete $self->{record};
313    
314            # read XRFMFB abd XRFMFP
315            read($self->{'fileXRF'}, $buff, 4);
316            my $pointer=unpack("l",$buff) || carp "pointer is null";
317    
318          print "XRFMFB: $XRFMFB XRFMFP: $XRFMFP\n" if ($self->{debug});          # check for logically deleted record
319            if ($pointer < 0) {
320                    print STDERR "## record $mfn is logically deleted\n" if ($self->{debug});
321                    $self->{deleted} = $mfn;
322    
323          # XXX fix this to be more readable!!                  return unless $self->{include_deleted};
         # e.g. (XRFMFB - 1) * 512 + XRFMFP  
324    
325          my $offset = $pointer;                  $pointer = abs($pointer);
         my $offset2=int($offset/2048)-1;  
         my $offset22=int($offset/4096);  
         my $offset3=$offset-($offset22*4096);  
         if ($offset3>512) {  
                 $offset3=$offset3-2048;  
326          }          }
         my $offset4=($offset2*512)+$offset3;  
327    
328          print "$offset - $offset2 - $offset3 - $offset4\n" if ($self->{debug});          my $XRFMFB = int($pointer/2048);
329            my $XRFMFP = $pointer - ($XRFMFB*2048);
330    
331            # (XRFMFB - 1) * 512 + XRFMFP
332            # why do i have to do XRFMFP % 1024 ?
333    
334            my $blk_off = (($XRFMFB - 1) * 512) + ($XRFMFP % 1024);
335    
336          close(fileXRF);          print STDERR "## pointer: $pointer XRFMFB: $XRFMFB XRFMFP: $XRFMFP offset: $blk_off\n" if ($self->{'debug'});
337    
338          # Get Record Information          # Get Record Information
339    
340          open(fileMST, $self->{isisdb}.".MST") || croak "can't open '$self->{isisdb}.MST': $!";          seek($self->{'fileMST'},$blk_off,0);
341    
342          seek(fileMST,$offset4,0);          read($self->{'fileMST'}, $buff, 4);
343            my $value=unpack("l",$buff);
344    
345          my $value=$self->Read32(\*fileMST);          print STDERR "## offset for rowid $value is $blk_off (blk $XRFMFB off $XRFMFP)\n" if ($self->{debug});
346    
347          if ($value!=$mfn) {          if ($value!=$mfn) {
348  print ("Error: The MFN:".$mfn." is not found in MST(".$value.")");                      carp "Error: MFN ".$mfn." not found in MST(".$value.")";    
349                  return -1;      # XXX deleted record?                  #return;
350          }          }
351    
352  #       $MFRL=$self->Read16($fileMST);          read($self->{'fileMST'}, $buff, 14);
 #       $MFBWB=$self->Read32($fileMST);  
 #       $MFBWP=$self->Read16($fileMST);  
 #       $BASE=$self->Read16($fileMST);  
 #       $NVF=$self->Read16($fileMST);  
 #       $STATUS=$self->Read16($fileMST);  
   
         my $buff;  
         read(fileMST, $buff, 14);  
353    
354          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);          my ($MFRL,$MFBWB,$MFBWP,$BASE,$NVF,$STATUS) = unpack("slssss", $buff);
355    
356          print "MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug});          print STDERR "## MFRL: $MFRL MFBWB: $MFBWB MFBWP: $MFBWP BASE: $BASE NVF: $NVF STATUS: $STATUS\n" if ($self->{debug});
357    
358            warn "MFRL $MFRL is not even number" unless ($MFRL % 2 == 0);
359    
360            warn "BASE is not 18+6*NVF" unless ($BASE == 18 + 6 * $NVF);
361    
362          # Get Directory Format          # Get Directory Format
363    
# Line 277  print ("Error: The MFN:".$mfn." is not f Line 365  print ("Error: The MFN:".$mfn." is not f
365          my @FieldLEN;          my @FieldLEN;
366          my @FieldTAG;          my @FieldTAG;
367    
368          for (my $i = 0 ; $i < $NVF ; $i++) {          read($self->{'fileMST'}, $buff, 6 * $NVF);
369    
370  #               $TAG=$self->Read16($fileMST);          my $rec_len = 0;
 #               $POS=$self->Read16($fileMST);  
 #               $LEN=$self->Read16($fileMST);  
371    
372                  read(fileMST, $buff, 6);          for (my $i = 0 ; $i < $NVF ; $i++) {
373                  my ($TAG,$POS,$LEN) = unpack("sss", $buff);  
374                    my ($TAG,$POS,$LEN) = unpack("sss", substr($buff,$i * 6, 6));
375    
376                  print "TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});                  print STDERR "## TAG: $TAG POS: $POS LEN: $LEN\n" if ($self->{debug});
377    
378                  # The TAG does not exists in .FDT so we set it to 0.                  # The TAG does not exists in .FDT so we set it to 0.
379                  #                  #
# Line 301  print ("Error: The MFN:".$mfn." is not f Line 388  print ("Error: The MFN:".$mfn." is not f
388                  push @FieldTAG,$TAG;                  push @FieldTAG,$TAG;
389                  push @FieldPOS,$POS;                  push @FieldPOS,$POS;
390                  push @FieldLEN,$LEN;                  push @FieldLEN,$LEN;
391    
392                    $rec_len += $LEN;
393          }          }
394    
395          # Get Variable Fields          # Get Variable Fields
396    
397          delete $self->{record};          read($self->{'fileMST'},$buff,$rec_len);
398    
399            print STDERR "## rec_len: $rec_len poc: ",tell($self->{'fileMST'})."\n" if ($self->{debug});
400    
401          for (my $i = 0 ; $i < $NVF ; $i++) {          for (my $i = 0 ; $i < $NVF ; $i++) {
402                  my $rec;                  # skip zero-sized fields
403                  read(fileMST,$rec,$FieldLEN[$i]);                  next if ($FieldLEN[$i] == 0);
                 push @{$self->{record}->{$FieldTAG[$i]}}, $rec;  
         }  
         close(fileMST);  
404    
405          # The record is marked for deletion                  push @{$self->{record}->{$FieldTAG[$i]}}, substr($buff,$FieldPOS[$i],$FieldLEN[$i]);
         if ($STATUS==1) {  
                 return -1;  
406          }          }
407    
408          print Dumper($self) if ($self->{debug});          $self->{'current_mfn'} = $mfn;
409    
410            print STDERR Dumper($self),"\n" if ($self->{debug});
411    
412          return $self->{'record'};          return $self->{'record'};
413  }  }
414    
415  =head2 to_ascii  =head2 to_ascii
416    
417  Dump ascii output of selected MFN  Dump ASCII output of record with specified MFN
418    
419    print $isis->to_ascii(55);    print $isis->to_ascii(42);
420    
421    It outputs something like this:
422    
423      210   ^aNew York^cNew York University press^dcop. 1988
424      990   2140
425      990   88
426      990   HAY
427    
428    If C<read_fdt> is specified when calling C<new> it will display field names
429    from C<.FDT> file instead of numeric tags.
430    
431  =cut  =cut
432    
# Line 337  sub to_ascii { Line 435  sub to_ascii {
435    
436          my $mfn = shift || croak "need MFN";          my $mfn = shift || croak "need MFN";
437    
438          my $rec = $self->GetMFN($mfn);          my $rec = $self->fetch($mfn);
439    
440          my $out = "0\t$mfn";          my $out = "0\t$mfn";
441    
442          foreach my $f (sort keys %{$rec}) {          foreach my $f (sort keys %{$rec}) {
443                  $out .= "\n$f\t".join("\n$f\t",@{$self->{record}->{$f}});                  my $fn = $self->tag_name($f);
444                    $out .= "\n$fn\t".join("\n$fn\t",@{$self->{record}->{$f}});
445          }          }
446    
447          $out .= "\n";          $out .= "\n";
# Line 350  sub to_ascii { Line 449  sub to_ascii {
449          return $out;          return $out;
450  }  }
451    
452  ################# old cruft which is not ported from php to perl  =head2 to_hash
   
 =begin php  
   
   # Load the dictionary from the $db.L0x files.  
   # Not usefull Yet  
     
   sub LoadDictionary()  
   {  
     $fileL01=fopen($self->{isisdb}.".L01","r");  
     rewind($fileL01);    
   
     do  
     {  
   
       $POS=$self->Read32($fileL01);  
       $OCK=$self->Read16($fileL01);  
       $IT=$self->Read16($fileL01);  
       $PS=$self->Read32($fileL01);  
 print "<br>PS:".$PS." ".$self->{ORDF}->{1}." ";  
       for ($i=0;$i<$OCK;$i++)  
       {  
         $KEY=fread($fileL01,10);  
         
         print $KEY." ### ";  
   
         $INFO1=$self->Read32($fileL01);  
         $INFO2=$self->Read32($fileL01);  
   
         #L01Key->{$key}=array($INFO1,$INFO2);  
       }  
       
       rewind($fileL01);  
       $offset=($PS-1)*(12+$self->{ORDF}->{1}*18*2);  
       fseek($fileL01,$offset);  
   
     } While (!feof($fileL01));  
   
     fclose($fileL01);  
   }  
   
   # self function search through the tree and returns an array of pointers to IFP  
   # The function must be recursive  
453    
454    sub SearchTree($search,$fileNB,$PUNT)  Read record with specified MFN and convert it to hash
   {        
       $offset=(($PUNT-1)*(8+2*$self->{ORDN}->{1}*14));  
   
         rewind($fileNB1);  
   
         fseek($fileNB,$offset);  
   
         $POS=$self->Read32($fileNB);  
         $OCK=$self->Read16($fileNB);  
         $IT=$self->Read16($fileNB);  
   
 #print "<br>".$POS." - ".$OCK." - ".$IT;  
   
         $OLDPUNT=$POS;  
         $j=0;  
         for ($i=0;$i<$OCK;$i++)  
         {  
           $KEY=fread($fileNB,10);  
         
           $PUNT=$self->Read32($fileNB);  
   
 #print " ## ".chop($KEY)."(".$PUNT."-".$OLDPUNT.") ## ";  
   
           If (strcmp($search,chop($KEY))<0)  
           {  
             break;  
           }  
           $OLDPUNT=$PUNT;    
         }          
 #print $OLDPUNT;  
         Return $OLDPUNT;  
   }  
455    
456    # Search ISIS for record containing search    my $hash = $isis->to_hash($mfn);
   # Return a sorted array of MFN  
457    
458    sub Search($search)  It has ability to convert characters (using C<hash_filter> from ISIS
459    {  database before creating structures enabling character re-mapping or quick
460    fix-up of data.
461    
462    This function returns hash which is like this:
463    
464      $hash = {
465        '210' => [
466                   {
467                     'c' => 'New York University press',
468                     'a' => 'New York',
469                     'd' => 'cop. 1988'
470                   }
471                 ],
472        '990' => [
473                   '2140',
474                   '88',
475                   'HAY'
476                 ],
477      };
478    
479    You can later use that hash to produce any output from ISIS data.
480    
481    If database is created using IsisMarc, it will also have to special fields
482    which will be used for identifiers, C<i1> and C<i2> like this:
483    
484      '200' => [
485                 {
486                   'i1' => '1',
487                   'i2' => ' '
488                   'a' => 'Goa',
489                   'f' => 'Valdo D\'Arienzo',
490                   'e' => 'tipografie e tipografi nel XVI secolo',
491                 }
492               ],
493    
494    $search=strtoupper($search);  This method will also create additional field C<000> with MFN.
 #print "Searching....".$search." - ".$self->{POSRX}->{1}."<br>";  
     # first search .x01  
       
   
     # Search in .N01    
   
   
     $fileN01=fopen($self->{isisdb}.".N01","r");  
     $offset=(($self->{POSRX}->{1}-1)*(8+2*$self->{ORDN}->{1}*14));  
   
       do  
       {  
         rewind($fileN01);  
   
         fseek($fileN01,$offset);  
   
         $POS=$self->Read32($fileN01);  
         $OCK=$self->Read16($fileN01);  
         $IT=$self->Read16($fileN01);  
   
 #print "<br>".$POS." - ".$OCK." - ".$IT;  
   
         $OLDPUNT=$POS;  
         for ($i=0;$i<$OCK;$i++)  
         {  
           $KEY=fread($fileN01,10);  
         
           $PUNT=$self->Read32($fileN01);  
   
 #print " ## ".chop($KEY)."(".$PUNT."-".$OLDPUNT.") ## ";  
   
           If (strcmp($search,chop($KEY))<0)  
           {  
             break;  
           }  
           $OLDPUNT=$PUNT;    
         }  
         $offset=(($OLDPUNT-1)*(8+2*$self->{ORDN}->{1}*14));        
       } while ($OLDPUNT>0);  
 #print $OLDPUNT;  
   
   
     fclose($fileN01);  
   
     # Now look for records in .L01 file  
     $fileL01=fopen($self->{isisdb}.".L01","r");  
     rewind($fileL01);  
   
     $offset=(-$OLDPUNT-1)*(12+$self->{ORDF}->{1}*18*2);  
     fseek($fileL01,$offset);  
   
     $POS=$self->Read32($fileL01);  
     $OCK=$self->Read16($fileL01);  
     $IT=$self->Read16($fileL01);  
     $PS=$self->Read32($fileL01);  
 #print "<br>POS:".$POS." ".$self->{ORDF}->{1}." ";  
     for ($i=0;$i<$OCK;$i++)  
     {  
       $KEY=fread($fileL01,10);  
         
 #print $KEY." ### ";  
   
       $INFO1=$self->Read32($fileL01);  
       $INFO2=$self->Read32($fileL01);  
   
       If (strcmp($search,chop($KEY))==0)  
       {  
         break;  
       }  
     }      
   
     fclose($fileL01);  
   
 #print $INFO1."--".$INFO2;  
   
     # Now look in .IFP for the MFN  
     $fileIFP=fopen($self->{isisdb}.".IFP","r");  
     rewind($fileIFP);  
     $offset=($INFO1-1)*512+($INFO2*4);  
     fseek($fileIFP,$offset);    
   
     $IFPBLK=$self->Read32($fileIFP);  
   
     $IFPNXTB=$self->Read32($fileIFP);  
     $IFPNXTP=$self->Read32($fileIFP);  
     $IFPTOTP=$self->Read32($fileIFP);  
     $IFPSEGP=$self->Read32($fileIFP);  
     $IFPSEGC=$self->Read32($fileIFP);  
   
   
 #print "<br>IFP:".$IFPBLK." # ".$IFPNXTB." - ".$IFPNXTP." - ".$IFPTOTP." - ".$IFPSEGP." - ".$IFPSEGC;  
   
     rewind($fileIFP);  
     $offset=($INFO1-1)*512+24+($INFO2*4);  
     fseek($fileIFP,$offset);      
       
     $j=24+($INFO2*4);  
     $k=0;  
     $l=1;  
     $OLDPMFN="";  
     for ($i=0;$i<$IFPSEGP;$i++)  
     {  
       $B1=$self->Read8($fileIFP);  
       $B2=$self->Read8($fileIFP);  
       $B3=$self->Read8($fileIFP);  
       $B4=$self->Read8($fileIFP);  
       $B5=$self->Read8($fileIFP);  
       $B6=$self->Read8($fileIFP);  
       $B7=$self->Read8($fileIFP);  
       $B8=$self->Read8($fileIFP);  
   
       $PMFN=$B1*65536+$B2*256+$B3;  
       $PTAG=$B4*256+$B5;  
       $POCC=$B6;  
       $PCNT=$B7*256+$B8;  
   
       if ($OLDPMFN!=$PMFN)  
       {  
         if ($PMFN!=0)  
         {  
           $self->{MFNArray}->{$l}=$PMFN;  
           $OLDPMFN=$PMFN;  
           $l+=1;  
         }  
       }  
   
       $j=$j+8;  
 #print "<br>".$PMFN."-".$PTAG." - ".$POCC." - ".$PCNT;  
 #print "@@".$j."@@@@";  
       if ($j>=504)  
       {  
         if ($IFPNXTB==0 && $IFPNXTP==0)  
         {  
           $k=$k+1;  
           rewind($fileIFP);  
           $offset=($INFO1-1+$k)*512;    
           fseek($fileIFP,$offset);        
           $B=$self->Read32($fileIFP);  
 #print "<br>-".$B."-<br>";  
           $j=0;  
         } else  
         {  
           rewind($fileIFP);  
           $offset=($IFPNXTB-1)*512;    
           fseek($fileIFP,$offset);  
   
           $OLDIFPNXTB=$IFPNXTB;  
           $OLDIFPNXTP=$IFPNXTP;  
   
           $IFPBLK=$self->Read32($fileIFP);  
   
           $IFPNXTB=$self->Read32($fileIFP);  
           $IFPNXTP=$self->Read32($fileIFP);  
           $IFPTOTP=$self->Read32($fileIFP);  
           $IFPSEGP=$self->Read32($fileIFP);  
           $IFPSEGC=$self->Read32($fileIFP);  
   
           rewind($fileIFP);  
           $offset=($OLDIFPNXTB-1)*512+24+($OLDIFPNXTP*4);  
           fseek($fileIFP,$offset);      
       
           $j=24+($OLDIFPNXTP*4);  
           $k=0;  
           $j=0;  
         }  
       }  
   
     }      
     fclose($fileIFP);  
     return $l-1;  
   }  
495    
496  =cut  =cut
497    
498  #  sub to_hash {
 # XXX porting from php left-over:  
 #  
 # do I *REALLY* need those methods, or should I use  
 # $self->{something} directly?  
 #  
 # Probably direct usage is better!  
 #  
   
 sub GetFieldName {  
         my $self = shift;  
         return $self->{FieldName};  
 }  
   
 sub GetTagName {  
         my $self = shift;  
         return $self->{TagName};  
 }  
   
 sub GetFieldTag {  
499          my $self = shift;          my $self = shift;
         return $self->{FieldTAG};  
 }  
500    
501  sub GetNextMFN {          my $mfn = shift || confess "need mfn!";
         my $self = shift;  
         return $self->{NXTMFN};  
 }  
502    
503  sub GetMFNArray {          # init record to include MFN as field 000
504          my $self = shift;          my $rec = { '000' => [ $mfn ] };
         return $self->{MFNArray};  
 }  
 =begin php  
505    
506    sub Read32($fileNB)          my $row = $self->fetch($mfn);
   {  
     $B1=ord(fread($fileNB,1));  
     $B2=ord(fread($fileNB,1));  
     $B3=ord(fread($fileNB,1));  
     $B4=ord(fread($fileNB,1));  
   
     if ($B4<=128)  
     {  
       $value=$B1+$B2*256+$B3*65536+$B4*16777216;  
     } else  
     {  
       $value=$self->Not8($B1)+$self->Not8($B2)*256+$self->Not8($B3)*65536+$self->Not8($B4)*16777216;  
       $value=-($value+1);  
     }  
 #    print "(".$B1.",".$B2.",".$B3.",".$B4.":".$value.")";  
507    
508      return $value;            foreach my $k (keys %{$row}) {
509    }                  foreach my $l (@{$row->{$k}}) {
510    
511    sub Read24($fileNB)                          # filter output
512    {                          $l = $self->{'hash_filter'}->($l) if ($self->{'hash_filter'});
     $B1=ord(fread($fileNB,1));  
     $B2=ord(fread($fileNB,1));  
     $B3=ord(fread($fileNB,1));  
513    
514      $value=$B1+$B2*256+$B3*65536;                          my $val;
515    
516  #    print "(".$B1.",".$B2.",".$B3.":".$value.")";                          # has identifiers?
517                            ($val->{'i1'},$val->{'i2'}) = ($1,$2) if ($l =~ s/^([01 #])([01 #])\^/\^/);
518    
519      return $value;                            # has subfields?
520    }                          if ($l =~ m/\^/) {
521                                    foreach my $t (split(/\^/,$l)) {
522                                            next if (! $t);
523                                            $val->{substr($t,0,1)} = substr($t,1);
524                                    }
525                            } else {
526                                    $val = $l;
527                            }
528    
529    sub Read16($fileNB)                          push @{$rec->{$k}}, $val;
530    {                  }
531      $B1=ord(fread($fileNB,1));          }
     $B2=ord(fread($fileNB,1));  
532    
533      $value=$B1+$B2*256;          return $rec;
534  #    print "(".$B1.",".$B2.":".$value.")";  }
535    
536      return $value;    =head2 tag_name
   }  
537    
538    sub Read8($fileNB)  Return name of selected tag
   {  
     $B1=ord(fread($fileNB,1));  
539    
540      $value=$B1;   print $isis->tag_name('200');
 #    print "(".$value.")";  
541    
542      return $value;    =cut
   }  
543    
544    sub Not8($value)  sub tag_name {
545    {          my $self = shift;
546      $value=decbin($value);          my $tag = shift || return;
547      if (strlen($value)<8)          return $self->{'TagName'}->{$tag} || $tag;
     {  
       $buffer="";  
       for($i=0;$i<(8-strlen($value));$i++)  
       {  
         $buffer.="0";  
       }  
       $value=$buffer.$value;  
     }  
     $value=ereg_replace("0","3",$value);  
     $value=ereg_replace("1","0",$value);  
     $value=ereg_replace("3","1",$value);  
     $value=bindec($value);  
     return $value;  
   }  
548  }  }
549    
 =cut  
   
550  1;  1;
 __END__  
551    
552  =head1 BUGS  =head1 BUGS
553    
# Line 734  This module has been very lightly tested Line 560  This module has been very lightly tested
560          dpavlin@rot13.org          dpavlin@rot13.org
561          http://www.rot13.org/~dpavlin/          http://www.rot13.org/~dpavlin/
562    
563  This module is based heavily on code from LIBISIS.PHP - Library to read ISIS files V0.1.1  This module is based heavily on code from C<LIBISIS.PHP> library to read ISIS files V0.1.1
564  written in php and (c) 2000 Franck Martin - <franck@sopac.org> released under LGPL.  written in php and (c) 2000 Franck Martin <franck@sopac.org> and released under LGPL.
565    
566  =head1 COPYRIGHT  =head1 COPYRIGHT
567    
# Line 748  LICENSE file included with this module. Line 574  LICENSE file included with this module.
574    
575  =head1 SEE ALSO  =head1 SEE ALSO
576    
577  L<http://www.openisis.org|OpenIsis>, perl(1).  OpenIsis web site L<http://www.openisis.org>
578    
579    perl4lib site L<http://perl4lib.perl.org>
580    
Legend:
Removed from v.3  
changed lines
  Added in v.25
  ViewVC Help
Powered by ViewVC 1.1.26